It’s often best not to look a gift horse in the mouth—and it’s hard not to view a recent article in the New York Times offering a skeptical view of the recently passed federal infrastructure bill as a gift. Given how much the Biden administration has touted the infrastructure plan as a historic accomplishment, the article represents a welcome break from the Times’s usual Democratic Party boosterism. Unfortunately, the article—by Ralph Vartabedian, a former reporter at the Los Angeles Times who wrote frequently and skeptically about California’s high-speed rail project—is disappointing. It mis-frames the central problem with American infrastructure, ignores a far more informative international perspective, and ultimately overlooks several ways to fix the incentives under which infrastructure planners operate.
In Vartabedian’s framing, the main problem with American infrastructure projects is cost overruns. He offers several examples of American projects that ended up costing far more than their initial estimates. The expert that Vartabedian most quotes, meanwhile, is Bent Flyvbjerg, a Danish professor now at the University of Oxford. Flyvbjerg’s research, the most influential of which is now more than a decade old, centers on the causes of cost overruns in all forms of “megaprojects.” In 2003, the professor compiled a large international dataset for an influential paper finding that cost overruns on large infrastructure projects are usual, with average sizes ranging from 20 percent for road projects to 45 percent for rail. Flyvbjerg has proposed some common reasons, such as “strategic misrepresentation”: the leaders of megaprojects offer excessively optimistic cost estimates in order to get political support.
But overruns are the least of the problems with American infrastructure. Consider, for instance, the Second Avenue Subway in New York, projected to cost $3.8 billion in 2004 and completed for about $4.5 billion. Even if the finished project had come in at $3.8 billion, it would still have been twice as expensive as almost any other subway project in the world. Most of the case studies of cost overruns in Flyvbjerg’s work come from the English-speaking world, but several more come from Scandinavia, and Flyvbjerg usually discusses these side-by-side as examples of the same universal problems. But the most salient fact about Scandinavia and the Anglosphere is their difference, not similarity: English-speaking nations have, without exception, the world’s most expensive infrastructure, while Scandinavia has some of the world’s cheapest. Even projects that had severe cost overruns, such as the City Circle Line in Copenhagen, are still inexpensive by world standards.
It’s better to view cost overruns merely as one part of the general problem of expensive infrastructure and ask why the United States is more vulnerable to this problem than elsewhere—an avenue of investigation that typically points to specific aspects of the American political environment. To take one example, a recent article by Flyvbjerg in the Harvard Business Review argued that large projects should use modular designs, made up of small reproducible components—and offered as an example an extension of the Madrid Metro, whose manager Flyvbjerg praises for “decid[ing] that no signature architecture would be used in the stations.” But the general advantages of modular design in all spheres of construction have been known for quite a while. Why, then, do some managers, but not others, choose cheap, uniform designs?
In the United States, one clear reason is a structural political incentive that is much weaker in other nations. New rail projects in the United States must run through endless community meetings where residents—often rich retirees who do not commute by public transit and have no work or childcare obligations that would keep them away from hours-long meetings—can demand additional spending to address local aesthetic concerns. Such a process was responsible, for example, for severe cost inflation on Boston’s Green Line Extension, according to preliminary research by the Transit Costs Project, a research team at New York University that takes an explicitly comparative international approach. Vartabedian’s only gesture at international comparisons, citing a professor at UC Berkeley, is extremely understated: “In some cases, U.S. construction costs are higher than those in Western Europe and democratic nations in Asia.”
This runaway localism is also responsible for much of the excessive cost of California High-Speed Rail, which receives extensive mention in Vartabedian’s article. The estimated unit costs for California High-Speed Rail—that is, the costs per mile of track, viaduct, tunnel, or retaining wall—were actually in line with international norms; excess costs stemmed largely from overengineering and poor route choices with traceable political causes. The most egregious example of this, the crossing of the Tehachapi Mountains from Los Angeles to the Central Valley, again stemmed from excessive interference of local interests. Likely at the behest of Los Angeles County officials who wanted to encourage suburban growth, the California High-Speed Rail Authority designed a long detour through the exurb of Palmdale rather than following the natural direct route to the north-northwest along Interstate 5. The authority justified its decision with an analysis filled with bizarre assumptions—asserting, for example, that the more direct route would require several miles of extra tunnels to avoid even minor impacts on a planned housing development in a key location.
A final detail of Vartabedian’s article is almost comical: an uncritical quotation of Ron Tutor, head of Tutor Perini, an infrastructure contractor highly active on the West Coast. Tutor rides to the defense of cost overruns: “All the major projects have cost and schedule issues,” he says. “The truth is these are very high-risk and difficult projects. Conditions change. It is impossible to estimate it accurately. That is naïve.”
Tutor would know, and it’s astonishing that Vartabedian didn’t mention his background. Tutor Perini is notorious in California and elsewhere for submitting low bids on contracts by reading them in bad faith to identify potential ambiguities, even if the intended meaning is obvious, then demanding payments for expensive “change orders” to do work in the way that the contract authors had obviously intended. According to a 2014 article in the Seattle Times, while Tutor Perini was building an expensive road tunnel in central Seattle, one lawyer who had litigated against the firm “said that much of Tutor’s business strategy appears to involve spending a lot of time working with attorneys before submitting a project bid, identifying areas where project documents could be considered ambiguous.” A 2012 article in the Hollywood Reporter repeats the allegations and adds, “Through the years, newspaper stories have suggested that campaign contributions by Tutor and his company have influenced the votes of officials who have approved change orders or new projects for Tutor Perini.” A Tutor Perini subsidiary has been accused of similar behavior in Pennsylvania.
The contractor’s ability to get away with such antics, of course, reflects other weaknesses of American politics. Alon Levy, one of the chief researchers on NYU’s Transit Costs Project, theorizes that bad contractor oversights result from a general devaluation of the civil service. Managerial positions in American transit and infrastructure agencies are often severely understaffed, to the point that the agencies cannot keep tabs on their contractors (even as blue-collar jobs are severely featherbedded), and the low pay compared with private-sector work, often backloaded into a pension with a vesting time of decades, makes finding high-quality, ambitious employees difficult. The American civil service is also not trusted with the same autonomy that civil services in other nations get. For instance, many jurisdictions—including, in most cases, California—have strict lowest-bidder rules for awarding contracts and prohibitively complex processes for disbarring bad contractors from future bids, a combination that enables bad actors such as Tutor Perini. In nations such as Spain, a world leader for low infrastructure costs, contracts are awarded on a combination of price and technical merit of contractors’ proposals, the latter quality assessed by the civil service.
It’s hard to fault Vartabedian too much. Many of the assumptions he makes are commonplace, and it’s heartening to see a paper like the Times entertain the notion that infrastructure bills might be judged on other criteria than dollars spent and jobs created. But his article pays mere lip service to the most important fact of America’s infrastructure problems: that most other developed nations don’t have them in anything near the same severity. Only a comparative international perspective can shed light on their causes, and overreliance on work such as Flyvbjerg’s that treats them mainly as instances of a universal pathology can obscure more than it enlightens.
Photo by Gary Coronado / Los Angeles Times via Getty Images